The Human KLK8 (Neuropsin/Ovasin) Gene: Identification of Two Novel Splice Variants and Its Prognostic Value in Ovarian Cancer

806 Vol. 7, 806–811, April 2001 Clinical Cancer Research

Angeliki Magklara, Andreas Scorilas, KLK8 encodes for a predicted secreted protein, its detection Dionyssios Katsaros, Marco Massobrio, in serum may aid in ovarian cancer diagnosis. George M. Yousef, Stefano Fracchioli, 1 Saverio Danese, and Eleftherios P. Diamandis INTRODUCTION Department of Pathology and Laboratory Medicine, Mount Sinai The human kallikrein gene family is a subfamily of serine Hospital and Department of Laboratory Medicine and Pathobiology, proteases, located at the chromosomal locus 19q13.3–q13.4. University of Toronto, Toronto, Ontario, M5G 1X5 Canada [A. M., A. S., G. M. Y., E. P. D.], and Departments of Obstetrics and Until recently, this family was known to include only three Gynecology, Gynecologic Oncology Unit [D. K., M. M., S. F.] and members: the pancreatic/renal kallikrein gene (KLK1), the hu- Gynecology, S. Anna Hospital [S. D.], University of Turin, Turin man glandular kallikrein 2 gene (KLK2), and prostate-specific 10126, Italy antigen (KLK3). In the past few years, another 11 kallikrein-like genes were discovered (reviewed in Ref. 1). Neuropsin is one of ABSTRACT these genes that maps to this locus (1, 2). According to the KLK8 (neuropsin/ovasin) is a new member of the human approved human kallikrein gene nomenclature, neuropsin is also kallikrein gene family, which consists of enzymes with serine known as KLK8 (3). protease enzymatic activity. Recent reports have implicated Neuropsin is a well-characterized, brain-related serine KLK8 in ovarian cancer. KLK8 may have potential clinical protease in mouse, where it has been shown to play an value for disease diagnosis or prognosis and it may also be a important role in neural plasticity (4), as well as in embryonic useful therapeutic target. cell differentiation (5). Recently, the human homologue of Purpose: We undertook this study to evaluate the prog- mouse neuropsin was cloned and found to be highly ex- nostic value of KLK8 in ovarian carcinoma by examining its pressed in brain, as well as in skin (6). Analysis of the expression in ovarian tumors. genomic organization of the human kallikrein 8 gene (KLK8) Experimental Design: The KLK8 gene was analyzed by revealed that it is composed of six exons and five introns, the reverse transcription-PCR and direct sequencing in several first exon being non-coding (6). The cDNA has a single open human normal tissues. Subsequently, its expression was reading frame of 780 bp encoding for a predicted 260-amino studied in a set of ovarian tumors, and statistical analysis acid protein. Two alternatively spliced forms of KLK8 have was performed. been identified so far (7, 8). The first (type 2 neuropsin) has Results: We have identified two novel mRNA splice an insert of 45 amino acids between the leader (pre)peptide variants of the KLK8 gene, which are abundantly expressed and the proenzyme (pro)peptide of the regular protein, be- in many tissues. These new variants were named KLK8 type cause of the presence of a second in-frame splice site in exon 3 and type 4. Study of the expression of the KLK8 gene and its spliced variants in ovarian tumors indicated that the new 3, which gives rise to a different mRNA form (7). The second variants were expressed very frequently and that full-length mRNA variant, known as tumor-associated differentially ex- KLK8 expression is an independent and favorable prognos- pressed gene-14, contains an additional new sequence of 491 Ј tic marker for ovarian cancer. Patients with higher KLK8 bp of 5 untranslated region; also, the nucleotides preceding expression in the tumor have lower grade disease, lower the poly(A) tail in the 3Ј untranslated region are not homol- residual tumor left after surgery, live longer, and relapse ogous with the regular form of neuropsin, but this mRNA less frequently. In multivariate analysis, higher KLK8 ex- form encodes for an identical protein as the human neuropsin pression was significantly associated with longer disease-free gene (8). Neuropsin/tumor-associated differentially ex- survival. pressed gene-14 was found to be overexpressed in ovarian Conclusions: These results suggest that KLK8 is a novel, carcinoma (8). In addition, a GenBank submission (accession favorable prognostic marker in ovarian cancer. Because number AF095742) refers to KLK8 as ovasin and indicates that it is a potential marker for ovarian cancer. In this report, we examine the expression of KLK8 in ovarian tumors by reverse transcription-PCR. While examining Received 9/11/00; revised 12/28/00; accepted 1/2/01. the expression of the gene in normal tissues, however, we found, The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked along with the expected PCR band, two additional bands present advertisement in accordance with 18 U.S.C. Section 1734 solely to in all samples examined. Further analysis revealed that the two indicate this fact. bands represent novel, alternatively spliced forms of the KLK8 1 To whom requests for reprints should be addressed, at Department of gene. We here describe the prognostic value of the expression Pathology and Laboratory Medicine, Mount Sinai Hospital, 600 Uni- versity Avenue, Toronto, Ontario, M5G 1X5 Canada. Phone: (416) 586- level of KLK8 and its two novel splice variants in a group of 143 8443; Fax: (416) 586-8628; E-mail: [email protected]. patients with ovarian cancer.

MATERIALS AND METHODS Table 1 Relationship between KLK8 (regular form) status and other Analysis of the KLK8 Gene by RT-PCR2 and Direct variables in 143 patients with primary ovarian cancer Sequencing. Total RNA isolated from 26 different human No. of patients (%) ␮ tissues was purchased from Clontech, Palo Alto, CA. Two gof KLK8 KLK8 RNA were reverse-transcribed into first-strand cDNA using the Patients negative positive P Superscript preamplification system (Life Technologies, Inc., KLK8 type 3 Gaithersburg, MD). The final volume was 20 ␮l. PCR was Negative 97 65 (67.0) 32 (33.0) 0.018a performed in a reaction mixture containing 1 ␮l of cDNA, 10 Positive 46 21 (45.7) 25 (54.3) mM Tris-HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl , 100 ␮M KLK8 type 4 2 a deoxynucleotide triphosphates, and 2.5 units of HotStar Taq Negative 47 33 (70.2) 14 (29.8) 0.11 Positive 96 53 (55.2) 43 (44.8) DNA polymerase (Qiagen, Mississauga, Ontario, Canada) in an Stage Eppendorf thermal cycler. The primers used were described I 27 12 (44.4) 15 (55.6) previously by Yoshida et al. (6), and their sequences were II 10 5 (50.0) 5 (50.0) 0.17b 5Ј-GTG-ACC-CCG-CCC-CTG-GAT-T-3Ј (forward) and 5Ј- III 90 59 (65.6) 31 (34.4) Ј IV 11 8 (72.7) 3 (27.3) GGG-AGA-TCT-AGT-GCT-TAT-CCT-3 (reverse). The cy- xc 5 cling conditions were 95°C for 15 min to activate the HotStar Grade Taq polymerase, followed by 40 cycles of 94°C for 30 s, 62°C GBd 10 2 (20.0) 8 (80.0) for 30 s, 72°C for 1 min, and a final extension step at 72°C for G1 17 8 (47.1) 9 (52.9) 0.007b 10 min. Equal amounts of PCR products were electrophoresed G2 23 12 (52.2) 11 (47.8) G3 89 62 (69.7) 27 (30.3) on 1% agarose gels and visualized by ethidium bromide stain- x4 ing. The three visible bands were cut out and purified using the Histotype Gel Purification kit (Qiagen), and the purified DNA was directly Serous 63 37 (58.7) 26 (41.3) b sequenced. To ensure that the first (non-coding) exon is in- Endometrioid 25 13 (52.0) 12 (48.0) 0.32 Undifferentiated 22 17 (77.3) 5 (22.7) cluded in the variant mRNAs, we constructed a new forward Others 31 18 (58.1) 13 (41.9) primer (5Ј-GTT-CCC-AGA-AGC-TCC-CCA-G-3Ј), binding at x2 the beginning of exon 1, and performed new PCR analysis using Residual tumor (cm) the aforementioned reverse primer. 0 56 27 (48.2) 29 (51.8) b Patients with Ovarian Cancer. Included in this study 1–2 27 16 (59.3) 11 (40.7) 0.044 Ͼ2 53 38 (71.7) 15 (28.3) were 143 consecutive patients with epithelial ovarian carci- x7 noma, with ages ranging from 25 to 80 with a median of 58 Menopause years. Patients underwent surgery and treatment for primary Pre/peri 49 27 (55.1) 22 (44.9) 0.47a ovarian carcinoma at the Department of Obstetrics and Gyne- Post 94 59 (62.8) 35 (37.2) Response to CTXe cology, Gynecological Oncology Unit, University of Turin, NC/PD 15 10 (66.7) 5 (33.3) 0.59a between 1991 and 1999. Follow-up information (median fol- CR/PR 118 69 (58.5) 49 (41.5) low-up period, 48 months) was available for 135 patients, NE 10 among whom 81 (60%) had relapsed and 57 (42%) died. Pa- a ␹2 test. tients were staged according to the criteria of the International b Fisher’s exact test. Federation of Gynecology and Obstetrics (9); information was c x, unknown status. d available for 138 of them (Table 1). Specifically, 27 (20%) GB, borderline grade. e CTX, chemotherapy; NC, no change; PD, progressive disease; patients had stage I disease, 10 (7%) had stage II, 90 (65%) had CR, complete response; PR, partial response; NE, not evaluated. stage III, and 11 (8%) had stage IV disease. All patients were treated with chemotherapy regimens containing platinum com- pounds, alone or in association with other drugs; grade 1 and stage I patients received no further treatment. Investigations undifferentiated, 11 were mucinous, 11 were clear cell, 7 were were performed in accordance with the Helsinki Declaration and Mullerian, and 2 were unclassified. In the data analysis, only were approved by the Institute of Obstetrics and Gynecology, serous, endometrioid, and undifferentiated cell carcinomas were University of Turin. considered. Because the numbers of the other histological types Ovarian Tumor Specimens and Extraction of Total were too small to be analyzed reliably, they were combined RNA. Each tumor had been histologically typed and graded together as one category called “others.” The size of the residual based on WHO criteria (10). Ten tumors (7%) were low poten- tumors ranged from 0 to 9 cm, with a median of 1.5 (Table 1). tial malignancies, 17 (12%) were Grade 1, 23 (17%) were Grade Tumor specimens were stored in liquid nitrogen immediately 2, and 89 (64%) were Grade 3. With respect to histological type, after frozen section and shipped at Ϫ80°C. Tumor tissue (50– 63 tumors were serous papillary, 25 were endometrioid, 22 were 100 mg) was pulverized on dry ice, followed by total RNA extraction using the TRIzol method (Life Technologies, Inc.). RT-PCR was performed as described above. To ensure integrity of RNA and efficiency of the reverse transcription, the house- 2 The abbreviations used are: RT-PCR, reverse transcription-PCR; PFS, keeping gene actin was also analyzed by PCR. progression-free survival; OS, overall survival; EST, expressed se- Assessment of KLK8 Expression. Agarose gel images quence tag; HR, hazard ratio; PSA, prostate-specific antigen. were scored by two independent investigators for high (positive)

Fig. 1 Tissue expression of KLK8 by RT- PCR. The expected PCR band (regular type) is 850 bp, type 3 is 450 bp, and type 4 is 300 bp long. There was no DNA template used in the PCR negative control. Actin was used as internal control for the integrity of the mRNA.

or low/undetectable levels of the three types of KLK8 (regular form, type 3 and type 4; Fig. 1). Type 2 KLK8 mRNA (described previously by Mitsui et al. (7)] was not detectable in any of the tumors. Concordance of scoring between the two investigators was 95%. In the remaining few discrepant cases, the final scoring was decided by re-review of the images. Statistical Analysis. KLK8 expression at the mRNA level was classified as positive or negative, and associations between KLK8 status and other qualitative variables were analyzed using the ␹2 or the Fisher exact test, where appropriate. Cox proportional hazard regression model was developed to evaluate the association (i.e., the hazard ratio and its confidence interval) between the potential prognostic marker and progression-free or overall survival. This analysis was conducted at both univariate and multivariate levels. The multivariate model was adjusted for KLK8 expression in tumors, patient age, clin- Fig. 2 PCR analysis of six ovarian tissues. Lanes 1–3 are noncancer- ical stage, tumor grade, residual tumor size, and histological ous tissues with undetectable or very low levels of KLK8 (negative). The rest of them are cancerous tissues. The marker lane has seven visible type. Survival analyses were performed by constructing Kaplan- bands with lengths in bp of 1050, 850, 600, 500, 400, 300, and 200 bp. Meier PFS and OS curves for KLK8-positive and KLK8-nega- Actin was used as internal control for the integrity of the mRNA. tive patients, and the log-rank test was used to examine the differences between them. Statistical analyses were performed using SAS software (SAS Institute, Cary, NC). These results revealed that these two bands corresponded to RESULTS new, alternatively spliced variants; we named them KLK8 type Identification of Two New KLK8 mRNA Variants. Af- 3 and type 4, respectively (GenBank accession number ter PCR amplification of the KLK8 cDNA, we observed two AF251125). To further verify the presence of these variants in additional bands, present in most of the normal human tissues tissues, we conducted a homology search using the BLASTN examined as well as in ovarian tumors (Figs. 1 and 2). We algorithm (11) on the National Center for Biotechnology Infor- determined the nucleotide sequence of these three PCR prod- mation web server, against the human EST database (dbEST). ucts. The sequence of the expected band was identical to that We identified two partially sequenced EST clones. One clone reported previously for the KLK8 (human neuropsin) mRNA (AI656124) included exons 2, 5, and 6; the other one (6). The two new bands were also identical to the KLK8 mRNA, (AA320217) included exons 2 and 6. These findings further except that whole exons were deleted. Specifically, the middle support our proposal that the KLK8 mRNA is alternatively band (Fig. 1) was missing two exons (exons 3 and 4), whereas spliced, yielding two new variants, as we had found by RT-PCR. the lowest band was missing three exons (exons 3, 4, and 5). All known KLK8 variants are shown in Fig. 3.

Table 2 Univariate and multivariate analysis of KLK8 with regard to PFS and OS PFS OS HRa 95% CIb P HRa 95% CIb P Univariate analysis

KLK8 Negative 1.00 1.00 Positive 0.44 0.27–0.71 Ͻ0.001 0.55 0.31–0.96 0.038 Stage of disease (ordinal) 2.91 2.04–4.16 Ͻ0.001 3.46 2.19–5.45 Ͻ0.001 Grading (ordinal) 2.21 1.56–3.10 Ͻ0.001 2.49 1.57–3.95 Ͻ0.001 Residual tumor (ordinal) 1.25 1.17–1.32 Ͻ0.001 1.30 1.21–1.40 Ͻ0.001 Histological typec 1.44 0.92–2.24 0.11 1.41 0.82–2.39 0.21 Age 1.01 0.99–1.03 0.26 1.01 0.99–1.04 0.27 Multivariate analysis

KLK8 Negative 1.00 1.00 Positive 0.57 0.34–0.94 0.027 0.81 0.45–1.47 0.49 Stage of disease (ordinal) 1.82 1.24–2.67 0.002 2.01 1.22–3.27 0.005 Grading (ordinal) 1.48 0.99–2.21 0.051 1.59 0.93–2.74 0.089 Residual tumor (ordinal) 1.12 1.04–1.21 0.002 1.19 1.09–1.31 Ͻ0.001 Histological typec 1.25 0.79–1.98 0.34 0.96 0.55–1.65 0.88 Age 1.01 0.98–1.03 0.48 1.01 0.98–1.04 0.29 a HR estimated from Cox proportional hazard regression model. b Confidence interval of the estimated HR. c Endometrial, undifferentiated, and others versus serous.

Tissue Expression. High levels of KLK8 mRNA were strongest predictor for both PFS and OS, with a hazard ratio of found in brain, cerebellum, pancreas, and placenta, as reported approximately 3 and 3.5, respectively. The Cox regression previously (Ref. 7; Fig. 1). We also detected high expression of model showed that the relative risk of relapse was significantly the gene in the salivary gland, uterus, thymus, breast, testis, and lower (by 56%) in KLK8-expressing ovarian tumors. Similarly, kidney. In these tissues, we detected all three spliced variants KLK8 mRNA positivity indicated a 45% decrease in the relative (data not shown). Notably, there was no KLK8 expression in the risk of death for these patients. The Kaplan-Meier survival normal ovarian tissue (Fig. 1). curves (Fig. 4) also demonstrate that patients with KLK8-posi- KLK8 Expression as a Prognostic Marker for Ovarian tive tumors have substantially longer PFS (P Ͻ 0.001) and OS Carcinoma. Eighty-six patients (60%) showed overexpres- (P ϭ 0.034) than patients with KLK8-negative tumors. sion of the regular form of KLK8, and 57 of them (40%) had When all variables were included in the Cox model (mul- very low or undetectable levels. Table 1 presents the relation- tivariate analysis in Table 2), clinical staging and residual tumor ships between KLK8 expression and other clinical or patholog- status were found to be independent prognostic factors for PFS ical variables, including clinical stage, histological grade, his- and OS. KLK8 expression significantly added to the prognostic totype, residual tumor after surgery, menopausal status, and power in the same multivariate regression model for PFS (HR, response to chemotherapy. Of the 89 patients with poorly dif- 0.57; P ϭ 0.027) but not for OS. ferentiated (G3) tumors, ϳ70% of them were negative for KLK8 We have also performed association analysis between type expression, and only 30% were positive. On the other hand, 80% 3 and type 4 variant expression and various clinicopathological of the patients with low potential malignancies were KLK8 variables and found no statistically significant associations (data positive, and only 20% were negative. The association between not shown). KLK8 expression and tumor grade was statistically significant by the ␹2 test (P ϭ 0.007). The positivity rates for KLK8 expression/stage were 56% for Stage I disease, 50% for Stage II, DISCUSSION 34% for Stage III, and 27% for Stage IV (P ϭ 0.17). The In this study, we have isolated and characterized two new positivity rates for KLK8 expression were 52% in patients with variants of the mRNA encoding for KLK8 (neuropsin). The no residual tumor after surgery, 41% in patients with a residual existence of multiple splice variants of mRNA is frequent tumor of 1–2 cm, and 28% in patients with residual tumor Ͼ2 among kallikreins. Several distinct RNA species transcribed cm (P ϭ 0.044). KLK8 expression was not related to cancer from the KLK3/PSA gene have been described in the literature histotype, menopausal status, or response to chemotherapy (Ta- (12–15). A number of alternate transcripts have also been re- ble 1). ported for the KLK2 gene (16, 17). KLK12 (18) and KLK13 (19), The strength of the associations between each clinicopath- two newly identified kallikreins, are present in various spliced ological variable and PFS or OS are demonstrated in the uni- forms in most of the tissues examined. variate analysis in Table 2. The stage of the disease was the The novel alternatively spliced forms of KLK8 mRNA are

Fig. 3 Genomic organization of the different splice forms of the ,ء the KLK8 gene. Arrow and position of the start and the ter- mination codon, respectively. The first exon (exon 1)isun- translated. H, D, and S represent the approximate position of the catalytic amino acids within the exons. AA, amino acid. Type 2 neuropsin was described by Mit- sui et al. (7). For more information, see text.

formed by deletion of whole exons. The first variant, called type 3 KLK8, is 584 bp long and is missing exons 3 and 4; the second variant, called type 4 KLK8, is 450 bp long and is missing exons 3, 4, and 5. The two variant mRNAs encode for putative proteins of 119 and 32 amino acids, respectively, which are missing two or three amino acids of the catalytic triad of serine proteases (Fig. 3). The signal peptide, which is necessary for the secretion of the regular form and is located within the first 32 amino acids, is also incompletely encoded, indicating that these variants encode for putative proteins that are likely not secreted. Expres- sion analysis of KLK8 showed that the variant mRNAs are abundant in several tissues (Fig. 1). A previous report (8) suggested that KLK8 might play a role as a prognostic marker in ovarian cancer. To investigate the potential clinical value of KLK8 and its new variants, we studied their expression in a set of ovarian tumors. Statistical analysis showed that the two spliced variants had no prognostic value. However, the expression of the regular form of the gene was found to be inversely associated with tumor grade and residual tumor volume (Table 1). Therefore, high KLK8 expression is associated with small, less aggressive tumors. These results suggest that KLK8 is a favorable prognostic indicator for ovarian cancer. These observations are further corroborated by our finding that high KLK8 expression is an independent prognostic factor of improved PFS (Fig. 4). Nine of the 10 non-cancerous ovarian tissues examined showed no or very low levels of KLK8 expression. It seems that KLK8 is not expressed much in the normal ovarian tissue, but its expression increases in the first stages of tumorigenesis, and then it is progressively suppressed as the malignancy advances. Because there is no information available on the pathophysio- logical role of KLK8 in ovarian tissue, it would be difficult to formulate a hypothesis that could explain the way of regulation of gene expression and the mechanism by which KLK8 confers Fig. 4 Kaplan-Meier survival curves presenting the association be- a favorable prognostic outcome in ovarian cancer. tween KLK8 expression and PFS (A)orOS(B). There is now growing evidence that many kallikreins are related to malignancy. PSA is the most valuable clinical marker for diagnosis and management of prostate cancer (20). PSA is also recognized as a favorable prognostic factor for breast tumors (26), and KLK7 (stratum corneum chymotryptic enzyme, cancer (21, 22). Recent reports suggest that hK2 could be HSCCE) has been shown to be expressed at abnormally high another useful diagnostic marker for prostate cancer (23, 24), levels in ovarian cancer (27). All kallikreins are thought to have whereas KLK4 (prostase), another kallikrein gene, seems to be serine protease catalytic activity and may be able to activate also implicated with this disease (25). KLK13 is reported to be each other or other molecules (growth factors, cytokines, and down-regulated in breast cancer (19), KLK6 (zyme) was shown others) in a cascade of events associated with tumorigenesis. to be differentially expressed in primary breast and ovarian Our results suggest that KLK8 may be involved in such pro-

cesses. The finding that a protease like KLK8 is a favorable coding for a variant protein of prostate-specific antigen. Cancer Res., prognostic tumor marker is not new. Other proteases (e.g., 59: 2820–2824, 1999. pepsinogen C and PSA) have also been associated with favor- 15. Tanaka, T., Isono, T., Yoshiki, T., Yuasa, T., and Okada, Y. A novel able cancer prognosis (21, 22, 28). form of prostate-specific antigen transcript produced by alternative In summary, we here provide the first evidence that higher splicing. Cancer Res., 60: 56–59, 2000. KLK8 expression is associated with more favorable outcomes of 16. Riegman, P. H., Vlietstra, R. J., van der Korput, H. A., Romijn, J., and Trapman, J. Identification and androgen-regulated expression of patients with ovarian cancer. Because KLK8 encodes for a two major human glandular kallikrein-1 (hGK-1) mRNA species. Mol. predicted secreted protein, analysis of this molecule in serum Cell. Endocrinol., 76: 181–190, 1991. may aid in ovarian cancer diagnosis and prognosis. This possi- 17. Liu, X. F., Essand, M., Vasmatzis, G., Lee, B., and Pastan, I. bility is under investigation. Identification of three new alternate human kallikrein 2 transcripts: evidence of long transcript and alternative splicing. Biochem. Biophys. REFERENCES Res. Commun., 64: 833–839, 1999. 18. Yousef, G. M., Magklara, A., and Diamandis, E. P. KLK12 is a 1. Diamandis, E. P., Yousef, G. M., Luo, L. Y., Magklara, A., and novel serine protease and a new member of the human kallikrein gene Obiezu, C. V. The new human kallikrein gene family: implications in family–differential expression in breast cancer. Genomics, 69: 331– carcinogenesis. Trends Endocrinol. Metab., 11: 54–60, 2000. 341, 2000. 2. Yousef, G. M., and Diamandis, E. P. The expanded human kallikrein 19. Yousef, G. M., Chang, A., and Diamandis, E. P. Identification and gene family: locus characterization and molecular cloning of a new characterization of KLK-L4, a new kallikrein-like gene that appears to member, KLK-L3. Genomics, 65: 184–194, 2000. be down-regulated in breast cancer tissues. J. Biol. Chem., 275: 11891– 3. Diamandis, E. P., Yousef, G. M., Clements, J., Ashworth, L. K., 11898, 2000. Yoshida, S., Egelrud, T., Nelson, P. S., Shiosaka, S., Little, S., Lilja, H., Stenman, U. H., Rittenhouse, H. G., and Wain, H. New nomenclature 20. Diamandis, E. P. Prostate specific antigen–its usefulness in clinical for the human tissue kallikrein gene family. Clin. Chem., 46: 1855– medicine. Trends Endocrinol. Metab., 9: 310–316, 1998. 1858, 2000. 21. Yu, H., Giai, M., Diamandis, E. P., Katsaros, D., Sutherland, D. J., 4. Yoshida, S., and Shiosaka, S. Plasticity-related serine proteases in the Levesque, M. A., Roagna, R., Ponzone, R., and Sismondi, P. Prostate- brain. Int. J. Mol. Med., 3: 405–409, 1999. specific antigen is a new favorable prognostic indicator for women with breast cancer. Cancer Res., 55: 2104–2110, 1995. 5. Suzuki, J., Yoshida, S., Chen, Z. L., Momota, Y., Kato, K., Hirata, A., and Shiosaka, S. Ontogeny of neuropsin mRNA expression in the 22. Yu, H., Levesque, M. A., Clark, G. M., and Diamandis, E. P. mouse brain. Neurosci. Res., 23: 345–351, 1995. Prognostic value of prostate-specific antigen for women with breast 6. Yoshida, S., Taniguchi, M., Hirata, A., and Shiosaka, S. Sequence cancer: a large United States cohort study. Clin. Cancer Res., 4: 1489– analysis and expression of human neuropsin cDNA and gene. Gene 1497, 1998. (Amst.), 213: 9–16, 1998. 23. Magklara, A., Scorilas, A., Catalona, W. J., and Diamandis, E. P. 7. Mitsui, S., Tsuruoka, N., Yamashiro, K., Nakazato, H., and Yamagu- The combination of human glandular kallikrein (hK2) and free PSA chi, N. A novel form of human neuropsin, a brain-related serine prote- enhances the discrimination between prostate cancer and benign pros- ase, is generated by alternative splicing and is expressed preferentially tatic hyperplasia in patients with moderately elevated total PSA levels. in human adult brain. Eur. J. Biochem., 260: 627–634, 1999. Clin. Chem., 45: 1960–1966, 1999. 8. Underwood, L. J., Tanimoto, H., Wang, Y., Shigemasa, K., Parmley, 24. Kwiatkowski, M. K., Recker, F., Piironen, T., Pettersson, K., Otto, T. H., and O’Brien, T. J. Cloning of tumor-associated differentially T., Wernli, M., and Tscholl, R. In prostatism patients the ratio of human expressed gene-14, a novel serine protease overexpressed by ovarian glandular kallikrein to free PSA improves the discrimination between carcinoma. Cancer Res., 59: 4435–4439, 1999. prostate cancer and benign hyperplasia within the diagnostic “gray 9. Pettersson, F. Annual Report on the Treatment in Gynecological zone” of total PSA 4 to 10 ng/mL. Urology, 52: 360–365, 1998. Cancer, Vol. 22, pp. 83–102. Stockholm: International Federation of 25. Nelson, P. S., Gan, L., and Fergusson, C. Molecular cloning and Gynecology and Obstetrics, 1994. characterisation of prostase, an androgen-regulated serine protease with 10. Serov, S. F., and Sorbin, L. H. Histological Typing of Ovarian prostate-restricted expression. Proc. Natl. Acad. Sci. USA, 96: 3114– Tumors, Vol. 17, No. 9. Geneva: WHO, 1973. 3119, 1999. 11. Altschul, S. F., Madden, T. L., Scha¨ffer, A. A., Zhang, J., Zhang, Z., 26. Anisowicz, A., Sotiropoulou, G., Stenman, G., Mok, S. C., and Miller, W., and Lipman, D. J. Gapped BLAST and PSI-BLAST: a new Sager, R. A novel protease homolog differentially expressed in breast generation of protein database search programs. Nucleic Acids Res., 25: and ovarian cancer. Mol. Med., 2: 624–636, 1996. 3389–3402, 1997. 27. Tanimoto, H., Underwood, L. J., Shigemasa, K., Yan Yan, M. S., 12. Riegman, P. H., Klaassen, P., van der Korput, J. A., Romijn, J. C., Clarke, J., Parmley, T. H., and O’Brien, T. J. The stratum corneum and Trapman, J. Molecular cloning and characterization of novel pros- chymotryptic enzyme that mediates shedding and desquamation of skin tate antigen cDNAs. Biochem. Biophys. Res. Commun., 155: 181–188, cells is highly overexpressed in ovarian tumor cells. Cancer (Phila.), 86: 1988. 2074–2082, 1999. 13. Riegman, P. H., Vlietstra, R. J., van der Korput, J. A., Romijn, J. C., 28. Scorilas, A., Diamandis, E. P., Levesque, M. A., Diamandi-Papa- and Trapman, J. Characterization of the prostate-specific antigen gene: nastasiou, A., Khosravi, M. J., Giai, M., Ponzone, R., Roagna, R., a novel human kallikrein-like gene. Biochem. Biophys. Res. Commun., Sismondi, P., and Lopez-Otin, C. Immunoenzymatically determined 159: 95–102, 1989. pepsinogen C concentration in breast tumor cytosols: an independent 14. Heuze´, N., Olayat, S., Gutman, N., Zani, M., and Courty, Y. favorable prognostic factor in node-positive patients. Clin. Cancer Res., Molecular cloning and expression of an alternative hKLK3 transcript 5: 1778–1785, 1999.

Downloaded from clincancerres.aacrjournals.org on September 25, 2021. © 2001 American Association for Cancer Research. The Human KLK8 (Neuropsin/Ovasin) Gene: Identification of Two Novel Splice Variants and Its Prognostic Value in Ovarian Cancer

Angeliki Magklara, Andreas Scorilas, Dionyssios Katsaros, et al.

Clin Cancer Res 2001;7:806-811.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/7/4/806

Cited articles This article cites 26 articles, 10 of which you can access for free at: http://clincancerres.aacrjournals.org/content/7/4/806.full#ref-list-1

Citing articles This article has been cited by 21 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/7/4/806.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/7/4/806. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.